Two-way crosstalk between BER and c-NHEJ repair pathway is mediated by Pol-β and Ku70

Abstract

Multiple DNA repair pathways may be involved in the removal of the same DNA lesion caused by endogenous or exogenous agents. Although distinct DNA repair machinery fulfill overlapping roles in the repair of DNA lesions, the mechanisms coordinating different pathways have not been investigated in detail. Here, we show that Ku70, a core protein of nonhomologous end-joining (NHEJ) repair pathway, can directly interact with DNA polymerase-β (Pol-β), a central player in the DNA base excision repair (BER), and this physical complex not only promotes the polymerase activity of Pol-β and BER efficiency but also enhances the classic NHEJ repair. Moreover, we find that DNA damages caused by methyl methanesulfonate (MMS) or etoposide promote the formation of Ku70-Pol-β complexes at the repair foci. Furthermore, suppression of endogenous Ku70 expression by small interfering RNA reduces BER efficiency and leads to higher sensitivity to MMS and accumulation of the DNA strand breaks. Similarly, Pol-β knockdown impairs total-NHEJ capacity but only has a slight influence on alternative NHEJ. These results suggest that Pol-β and Ku70 coordinate 2-way crosstalk between the BER and NHEJ pathways.—Xia, W., Ci, S., Li, M., Wang, M., Dianov, G. L., Ma, Z., Li, L., Hua, K., Alagamuthu, K. K., Qing, L., Luo, L., Edick, A. M., Liu, L., Hu, Z., He, L., Pan, F., Guo, Z. Two-way crosstalk between BER and c-NHEJ repair pathway is mediated by Pol-β and Ku70.

DNA is continuously exposed to environmental factors and endogenous stresses that result in a variety of DNA lesions. In view of the plethora of the lesion types, no single-repair pathway can deal with all of them. Instead, evolution has molded a tapestry of sophisticated, interwoven DNA repair systems that, as a whole, cover most (if not all) of the insults inflicted on a cell’s vital genetic information (1). However, there is little direct evidence to clarify the intrinsic link between these repair pathways. Recently, various studies have suggested that some key proteins of 1 repair pathway may play important roles in another repair process. For example, it was shown that mismatch-repair protein MSH6 (mutS homolog 6) associates with Ku70 and positively regulates NHEJ, one of the DNA double-strand break (DSB) repair pathways (2). Besides, another mismatch-repair protein mutL homolog 1 (MLH1) can modulate base excision repair (BER) and enhance the response to alkylation DNA damage (3). So far, a growing number of studies have been conducted to address the mechanisms of the crosstalk between different DNA repair pathways during DNA damage response. Because single-strand break (SSB) and DSB are typical DNA damages usually found in cells, it is extremely important to seek the mutual effect of SSB and DSB repair pathways. It will not only be helpful to understand the detailed mechanism of DNA damage response but will also have implications for the new strategy of cancer chemotherapy.

BER is an evolutionarily conserved pathway from bacteria to humans and is responsible for repairing most of the DNA damage, including lesions induced by oxidation, alkylation, and SSBs (4). During higher eukaryotic genome maintenance, BER functions in removal and repair of at least 20,000 DNA lesions per cell per day (5). BER is initiated by a DNA glycosylase that recognizes and removes the damaged base, leaving an abasic site that is further processed by short-patch BER (SP-BER) or long-patch BER (LP-BER) that largely uses different proteins to accomplish BER (6, 7). DNA polymerase-β (Pol-β) is a key enzyme implicated in BER pathway, where it mainly repairs SSBs (8, 9). It comprises a catalytic domain required for the polymerase activity and an 8-kDa N-terminal 5′-deoxyribosephosphate lyase (dRP) domain. Pol-β deficiency impairs BER efficiency and causes cells to be more sensitive to alkylating or oxidative agents (10).

Improper SSB repair may result in DSBs in replicating DNA. The repair of DSBs in mammalian cells is mainly accomplished by homologous recombination (HR) and nonhomologous end-joining (NHEJ). HR requires a region of extensive homology, operates during late S and G₂ phases of the cell cycle, and is accurate. In contrast, NHEJ is active at all stages of the cell cycle, does not rely on a template, and is error prone (11–13). Despite its error-prone nature, NHEJ is the predominant form of DSB repair in human somatic cells. Recent evidence suggests the existence of at least 2 subpathways, classic NHEJ (c-NHEJ) and alternative NHEJ (alt-NHEJ) (14). The major factors involved in c-NHEJ are the DNA-dependent protein kinase (DNA-PK) complex, which includes the Ku proteins (Ku70/Ku80 heterodimer), the catalytic subunit of DNA-PK (DNA-PKcs), and the XRCC4/ligase IV complex. The Ku proteins initiate NHEJ by sensing and binding to the DSBs. Subsequently, Ku70/Ku80 recruits DNA-PK (DNA-PKcs) that spatially stabilizes 2 break ends. DNA-PKcs is activated and transmits repair signals by phosphorylation of downstream proteins and DNA-end processing proteins, including series of nucleases and polymerases that process the broken ends into a structure suitable for ligation. Finally, the ligase IV-XRCC4 complex performs ligation of the DNA ends (15–18). Previous work pointed out that 3 members of the Pol X family (Pol-μ, Pol-λ, and TdT) are implicated in NHEJ (19–21). However, the identity of the gap-filling polymerase involved in NHEJ is still under discussion.

In the present study, we report that Pol-β, a BER pathway protein, physically interacts with Ku70 (NHEJ pathway protein), and this association is enhanced by DNA damage. By using BER assay in vitro reconstituted with purified recombinant proteins, we demonstrate the stimulating effect of Ku70 on the polymerase activity of Pol-β and BER efficiency. Moreover, we show that Ku70 deficient cells have reduced BER activity and are sensitive to methyl methanesulfonate (MMS). Reciprocally, Pol-β also plays a role in the end-joining process. Knocking down Pol-β impairs total NHEJ but has no significant influence on alt-NHEJ, suggesting that Pol-β functions in the c-NHEJ. Our results provide evidence for the 2-way crosstalk between Ku70 and Pol-β that modulates DNA repair through BER and NHEJ.

MATERIALS AND METHODS

Cell culture and drug treatments

The human cervix adenocarcinoma cell line HeLa and human embryonic kidney cell line HEK293T were obtained from the American Type Culture Collection repository (Manassas, VA, USA). U2OS alt-NHEJ-EGFP cell line was from Dr. Jun Huang (Zhejiang University, Hangzhou, China). U2OS EJ5-GFP cell line was from Songbai Liu (Suzhou Vocational Health College, Jiangsu, China). These cells were cultured in DMEM supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin. To induce DNA breaks, exponentially growing cells were treated with etoposide (ETO; S1225, dissolved in DMSO; Selleck chemicals, Houston, TX, USA) or MMS (129925, diluted with PBS; MilliporeSigma, Burlington, MA, USA) at the indicated concentrations at 37°C for various times. Where indicated, HeLa cells were treated with 1 μM NU7441 (S2638, dissolved in DMSO; Selleck Chemicals) and 10 μM RI-1 (S8077, dissolved in DMSO; Selleck Chemicals) to block NHEJ and HR correspondingly.

Plasmid construction and preparation of recombinant proteins

Full-length Ku70 was obtained by amplifying the Ku70 gene from pCMV-Ku70-HA (purchased from Sinobiological, Beijing, China) using the following primers: 5′-CCGGCGGCCGCGATGTCAGGGTGGGAGTCATA-3′ (Ku70 forward) and 5′-CGCGGATCCGTCCTGGAAGTGCTTGGTGA-3′ (Ku70 reverse). Full-length Ku70 was cloned into pFlag-CMV4 and pET-28b plasmids. Human Pol-β cDNA was obtained by amplifying the Pol-β gene from a human liver cDNA library and cloned into pFLAG-CMV4, PLVX-IRES-Puro, and pET-28b plasmids. Recombinant proteins Ku70, Pol-β, apurinic endonuclease 1 (APE1), flap endonuclease 1 (FEN1), and proliferating cell nuclear antigen (PCNA) were expressed in Escherichia coli BL21DE and were purified using Ni²⁺ affinity chromatography (GE Healthcare, Waukesha, WI, USA). Uracil-DNA glycosylase (UDG), DNA ligase I (Lig I), and DNA ligase IIIα (Lig IIIα) were purchased from New England Biolabs (Ipswich, MA, USA), and Ku80 was purchased from Abnova (Taipei, Taiwan).

Total NHEJ and alt-NHEJ reporter assays

Quantification of intracellular NHEJ efficiency was performed according to previous studies (22, 23). Briefly, the U2OS EJ5-GFP or alt-NHEJ-EGFP cells were seeded into 6-well plates at 1 × 10⁶ cells/well; 10 pmol specific small interfering (siRNA)-targeted Pol-β or nontargeting siRNA was transfected into these cells. Twenty-four hours following transfection, 0.8 μg pCAGGS-I-SceI expression plasmid or an empty vector was transfected into these cells. Seventy-two hours after first transfection, cells were collected and processed for flow cytometry analysis; each experiment was repeated 3 times.

In vitro polymerase activity assay

The polymerase activity assay was performed in the reaction buffer containing 50 mM Tris-HCl (pH 8.0), 10 mM MgCl₂, 2 mM DTT, 20 mM NaCl, 10% glycerol mixed with 0.1 μM fluorescein (FAM)-labeled DNA substrate Pol-β-nick-FAM (see Supplemental Table S1 for details), 50 mM of each dATP, dGTP, deoxycytidine triphosphate, and deoxythymidine thriphosphate (Takara, Kyoto, Japan), and 10 ng Pol-β. Various amounts of Ku70 (0, 50 ng, 100 ng, 0.8 μg) or corresponding amount of BSA were added to the reconstituted reactions. Reactions were carried out at 37°C for 30 min in the dark, stopped by addition of the stop buffer containing 95% formamide, 20 mM EDTA, 0.05% bromophenol blue, and 0.05% xylene cyanol, and heated (95°C) for 10 min. Finally, 10 ul of the reaction mixture were separated by 15% PAGE containing 8 M urea and analyzed by Odyssey FC (Li-Cor, Lincoln, NE, USA).

Western blot analysis and immunoprecipitation assay

Cells were lysed with ice-cold RIPA lysis buffer (P0013C; Beyotime Biotechnology, Shanghai, China). Lysates were quantified using the Bradford assay, and equal amounts of proteins were then resolved on 12% SDS-PAGE gels. Proteins were transferred to PVDF membranes. The membranes were blocked for 2 h in 1 time PBS containing 5% fat-free milk at room temperature and then incubated with the indicated primary antibodies overnight at 4°C. After 2 h room temperature incubation with a suitable horseradish peroxidase–conjugated secondary antibody, the membranes were scanned with Tanon 4500 Imaging System (Tanon, Shanghai, China) and quantified with ImageJ software (National Institutes of Health, Bethesda, MD, USA).

For Immunoprecipitation assay, cells were lysed with ice-cold IP lysis buffer (P0013; Beyotime Biotechnology) containing protease inhibitor and treated with 100 mg/ml DNase1 (M0303S; New England Biolabs) for 25 min at 37°C. The lysates were incubated with anti-Flag M2 Beads (MilliporeSigma) or protein A+G agarose (Beyotime Biotechnology) and corresponding antibodies or control IgG overnight at 4°C with rotation, and the beads were washed 4–6 times with 1× PBS containing 0.2% Tween 20 and eluted with SDS loading buffer. Samples were separated by SDS-PAGE and analyzed by Western blot with the indicated antibodies. The bands were detected with Tanon 4500 and quantified with ImageJ software.

Immunofluorescence

Cells were seeded into 6-well plates with slides on the bottom and incubated overnight. After MMS or ETO treatment, cells were washed with 1 time PBS 3 times and fixed with 4% paraformaldehyde made fresh in 1× PBS for 10 min, permeabilized with 0.1–0.5% Triton X-100 for 10 min at room temperature, and blocked with 1% BSA in 1 time PBS for 1 h at room temperature. After blocking, cells were incubated with indicated primary antibodies overnight at 4°C. Subsequently, cells were washed with 1 time PBS with Tween 20 and then incubated with fluorescent secondary antibodies conjugated with FITC for 2 h at room temperature. Next, cells were stained with DAPI and visualized by fluorescence microscope (80I 10-1500X; Nikon, Tokyo, Japan), and the images were captured with a charge-coupled-device camera.

For colocalization analysis, fluorescent signals were acquired by Nikon confocal microscope system A1(Nikon Ti-E-A1R) equipped with a Nikon Eclipse Ti microscope and a diode laser for a 405-nm line, a sapphire solid laser for a 488-nm line, and a diode laser for a 640-nm line. Emission signals were detected by a photomultiplier tube (DU4) preceded by emission filters BP 450/50 nm, BP 515/30 nm, and BP 585/65 nm for DAPI, Alexa Fluor 488, and Alexa Fluor 594, respectively. A Plan Apo λ ×60 oil objective lens was used for the imaging. Laser scanning and image acquisition were performed with Nikon Imaging Software NIS Elements AR. The images obtained were then exported as tagged image file format files and further adjusted by ImageJ. Quantification of colocalization was analyzed by ImageJ colocalization finder plugin (24).

Reconstituted BER assay

The reconstituted BER assay was performed as described previously in Guo et al. (10). Briefly, all the BER reactions were carried out in 20 μl of reaction buffer containing 40 mM HEPES-KOH (pH 7.8), 70 mM KCl, 7 mM MgCl₂, 1 mM DTT, 0.5 mM EDTA, 2 mM ATP, 50 μM each of dATP, dTTP, dGTP, and dCTP, or 8 μM 2 μCi (α-[³²P])-dCTP. For SP-BER reconstitution, purified proteins, including UDG (8 ng), APE1 (10 ng), Lig IIIα (20 ng), Pol-β (2.5 ng), and various amounts of Ku70 (0, 50 ng, 100 ng, 0.8 μg) or a corresponding amount of BSA were mixed with the SP-BER substrate Pol-β-U and incubated for 30 min at 37°C. For LP-BER, Pol-β-F substrate was incubated with the mixture of APE1 (10 ng), Pol-β (8 ng), FEN1 (10 ng), PCNA (10 ng), Lig I (20 ng), and various amounts of Ku70 (0, 50 ng, 100 ng, 0.8 μg) for 30 min at 37°C. Both reactions were stopped with stop buffer and heated at 95°C for 10 min, and 10 μl reaction mixture was separated by 15% PAGE containing 8 M urea and imaged by Odyssey FC. For cell extract activity, substrates (Pol-β-U or Pol-β-F) were incubated with the indicated amount of the total cell extracts prepared from cells treated with Ku70 siRNA or scramble siRNA. All analyses of cell extract were visualized by autoradiography.

Alkaline Comet assay

Alkaline Comet assay was performed using a Comet Assay Kit (KGA240; Keygen Biotech, Nanjing, China) according to the manufacturer’s recommendations. Briefly, The HEK293T cells were grown overnight, treated with 1 mM MMS for 30 min to induce tail moments, and then harvested for Comet assay. To assess the values of the percentage of DNA in the tail, the cells stained with propidium iodide were visualized by fluorescence microscope and analyzed using Comet assay software project.

Measurement of intracellular NADPH

Quantification of intracellular NADPH level was performed according to a previous study (25). Briefly, HEK293T cells were seeded in 96-well plates at 5 × 10³ cells per well. After overnight incubation, cells were treated with MMS (at indicated concentrations) for 1 h. After MMS treatment, 10 μl Cell Counting Kit 8 (MedChemExpress, Monmouth Junction, NJ, USA) were added into each well and then incubated for 4 h to allow formazan dye production, which was measured on a microplate spectrophotometer (Spectra Max 340PC384; Sunnyvale, CA, USA) at 450 nm with 650 nm as the reference filter. The decrease in intracellular NADPH level was determined by comparing the absorbance of treated cells with that of the control. The means and se for each cell line were calculated from at least 3 independent experiments, each in triplicate.

Drug sensitivity assay

Sensitivity to MMS and ETO was determined by cell survival assay. Cells were seeded into a 96-well plate (1500/well), incubated (overnight, 37°C, 5% CO₂), and treated (12 h, 37°C) with 1 mM MMS or multiple dilutions of ETO. Cell Counting Kit 8 (E1CK-000208, 10 μl; EnoGene, Nanjing, China) was added to each well, and the cells were further incubated for 2 h at 37°C before the measurement of the absorbance at 450 nm. At least 5 replicas for each group were averaged. Data are expressed as the percentage of growth relative to untreated controls.

RESULTS

Ku70 directly interacts with Pol-β

In previous studies, we have demonstrated that cells with Pol-β defect are more sensitive to DNA-damaging agents such as MMS, which induces DNA base lesions (10, 26, 27). This is consistent with the essential role of Pol-β in BER. However, we found that Pol-β–deficient cells have similarly higher sensitivity to ETO, an established DBS inducer (Supplemental Fig. S1A, B), which indicates that Pol-β may have some role in DSB repair. This observation promoted us to hypothesize that Pol-β may have potential roles in DSB repair. In support of this hypothesis, Ku70, the key DSB repair protein, was identified as the candidate binding partner of Pol-β in Wilson’s study (28). However, the mechanism and the biologic significance of the interaction between Pol-β and Ku70 still remains unknown. In this study, we first overexpressed Flag-Pol-β in HEK293T cells and coimmunoprecipitated Flag-Pol-β by anti-Flag M2 beads. We found Ku70, the core protein of NHEJ, was among the major hits of mass spectrometry analysis (unpublished results). To further confirm Pol-β’s interaction with Ku70, we performed a series of immunoprecipitation assays. First, we coimmunoprecipitated endogenous Ku70 and Pol-β in HEK293T cells using Ku70 antibodies (Fig. 1A). In control experiments, normal rabbit IgG did not immunoprecipitate Ku70 or Pol-β, indicating that the coimmunoprecipitation of Pol-β with Ku70 was not due to nonspecific antibody binding. Ku80, a well-known protein interacting with Ku70, was also detected in this experiment (Fig. 1A). We also coimmunoprecipitated exogenous Ku70 and endogenous Pol-β in Flag-Ku70 overexpressing HEK293T cells (Supplemental Fig. S2A). Moreover, endogenous Ku70 can also be detected in HEK293T cells overexpressing Flag-Pol-β after immunoprecipitation with Flag antibody (Supplemental Fig. S2B). These data indicate that Pol-β and Ku70 may interact with each other directly or indirectly in cells.

Figure 1.

Ku70 interacts with Pol-β. A) Cell lysates were immunoprecipitated with antibodies against Ku70 or normal IgG. Associated proteins were analyzed by Western blot with indicated antibodies. B, C) Purified Ku70 and Pol-β were immunoprecipitated with indicated antibodies or normal IgG (as a control) and analyzed by Western blot.

Because Pol-β and Ku70 are chromatin-bound proteins, we considered the interaction between Pol-β and Ku70 might be mediated by DNA or other proteins. To exclude this possibility, we pretreated cell lysates with DNase I before all the immunoprecipitation assays above, and the additional in vitro binding assays were performed with purified recombinant Pol-β and Ku70 proteins. Our data revealed that Pol-β directly interacts with Ku70 without any bridging proteins or DNA (Fig. 1B, C). Importantly, it is well known that Ku70 and Ku80 form a Ku heterodimer, and our data also showed there is a bond between Ku70 and Ku80 (Fig. 1A and Supplemental Fig. S2A) and that Ku80 is also found in Flag-Pol-β immunoprecipitants (Supplemental Fig. S2B). To further check whether Pol-β could also directly interact with Ku80, the interaction of purified Ku80 and Pol-β was examined in an in vitro binding assay. As shown in Supplemental Fig. S2C, Ku80 could not be pulled down by an anti–Pol-β antibody, ruling out the possibility of a direct interaction between them. We thus concluded that only Pol-β and Ku70 proteins could directly interact in the cells.

DNA damage increases the interaction between Ku70 and Pol-β

Because both Pol-β and Ku70 play important roles in DNA repair via BER and NHEJ, respectively, we hypothesized that the association between Ku70 and Pol-β may be enhanced by the DNA damage to contribute to both BER and NHEJ. To test this, HEK293T cells stably expressing Flag-Pol-β were treated with alkylating agent MMS, which causes DNA base lesions that are repaired through the BER pathway (29). Cells were then lysed and subjected to immunoprecipitation with anti-Flag M2 beads. The results showed that the binding between Ku70 and Pol-β was clearly enhanced after MMS treatment (Fig. 2A). A similar result was obtained by the immunoprecipitation with anti-Ku70 (Supplemental Fig. S3A). Then, ETO, an established DSB inducer, was used in similar experiments. We found that the interaction between Pol-β and Ku70 was notably increased with ETO treatment (Fig. 2B). The reciprocal immunoprecipitation confirmed this result (Supplemental Fig. S3B). Moreover, with the increase of ETO concentration, the interaction was remarkably enhanced (Supplemental Fig. S3C). To further confirm the increased association of Pol-β and Ku70 in cells in response to DNA damage, we employed immunofluorescence assay. HEK293T cells treated with PBS, DMSO, or DNA-damaging agents (MMS and ETO) were stained with Pol-β–specific antibody (Fig. 2C, D, red) and Ku70 antibody (Fig. 2C, D, green), and the cell nuclei were stained with DAPI (Fig. 2C, D, blue). The red (Fig. 2C, D, Pol β) and green (Fig. 2C, D, Ku70) pixel overlap indicates the colocalization of Pol-β and Ku70. We found that with treatment with DNA-damaging agents, the overlap was significantly higher than after the control treatment. These data further support the interaction between Pol-β and Ku70 at the sites of DNA damage. We thus concluded that Pol-β and Ku70 form a physical complex in cells, and their interaction can be enhanced under DNA damage stress.

Figure 2.

The Ku70 and Pol-β interaction is enhanced in response to DNA damage. A) HEK293T cells stably expressing Flag-Pol-β were treated with 1 mM MMS or PBS for 30 min and left to recover for 1 h. Cells were lysed, and proteins were immunoprecipitated from the lysates using anti-Flag M2 Beads. The precipitates were then subjected to Western blot analysis. B) Similar to A, HEK293T cells stably expressing Flag-Pol-β were treated with 10 μM ETO or DMSO for 1.5 h and released for 1 h. Cell lysates were prepared and subjected to immunoprecipitation and Western blot analysis with the indicated antibodies. The numbers in the figure represent the relative gray values of the bands below (regard control treatment group as 1, quantified with ImageJ software). C, D) HEK293T cells were treated with 1 mM MMS (C, the bottom part of the panel) and 10 μM ETO (D, lower panel) or treated with control reagents (C, D, upper panel) for 30 min and left to recover for 1 h. Ku70 and Pol-β were detected using antibodies specific for Ku70 and Pol-β, which were visualized with Alexa Fluor 488– or 594–conjugated secondary antibodies, respectively, followed by confocal microscopy. Colocalization was visualized by using the ImageJ colocalization finder plugin. Colocalization of Ku70 (green) and Pol-β (red) in cells appears yellow in merged images. Quantification of the colocalization between Ku70 and Pol-β was performed by using the colocalization function of ImageJ from 150 cells. This experiment was repeated 3 times. Scale bars, 10 μm. ***P < 0.001 (2-sided Student’s t test).

Ku70 enhances BER efficiency by promoting polymerase activity of Pol-β

After establishing that the Pol-β and Ku70 interaction is enhanced with DNA damage, we subsequently sought to determine the functional and biologic significance of this interaction. Between the BER proteins identified, Pol-β has been demonstrated to be a key player in both SP-BER and LP-BER (30). Therefore, we asked whether Ku70 would play a role in the BER process through the interaction with Pol-β. To address this question, in vitro SP-BER and LP-BER assays were performed using purified recombinant BER proteins. Uracil (U)-containing substrate (Pol-β-U-FAM, Supplemental Table S1) and tetrahydrofuran (F)-containing substrate (Pol-β-F-FAM, Supplemental Table S1) labeled with FAM group were used for SP-BER and LP-BER, respectively. In the absence of Pol-β, the U or F lesions are cleaved by the concerted action of UDG and APE1, which results in a nicked DNA duplex and produces the 20mer intermediate product (Fig. 3A, B, lane 1). Once Pol-β was added, this intermediate structure was further processed to generate 21mer intermediates (not ligated) and a fully repaired 41mer product (Fig. 3A, B, lane 2). For SP-BER, the BER products increased with the increased amount of Ku70 added in the reaction (Fig. 3A, lanes 3, 4, and 5). Indeed, about 60% additional fully repaired product was observed in the reaction containing the highest amount of Ku70 (Fig. 3A, lane 5). BSA was added in the control reactions and had no effect on the efficiency of SP-BER (Fig. 3A, lanes 6, 7, and 8). Additionally, without Pol-β, Ku70 itself failed to stimulate the repair process (Fig. 3A, lane 9). Similarly, Ku70 also obviously increased the efficiency of LP-BER (Fig. 3B, lanes 3, 4, and 5), whereas BSA failed to do so (Fig. 3B, lanes 6, 7, and 8).

Figure 3.

Ku70 promotes BER capacity by stimulating Pol-β activity. A) SP-BER was reconstituted with purified recombinant proteins and FAM-labeled U-containing oligonucleotide duplex substrate. Ku70 protein (0, 50, 100 ng, 0.8 μg) or similar amount of BSA were added to reconstituted reactions. B) LP-BER was reconstituted with purified recombinant proteins and FAM-labeled F-containing oligonucleotide duplex substrate. Ku70 protein (0, 50, 100 ng, 0.8 μg) or similar amount of BSA were added to the reconstituted reactions. C) Polymerase activity assay was performed using 20 ng purified Pol-β and FAM-labeled oligonucleotide duplex substrate. Where indicated, Ku70 protein (0, 50, 100 ng, 0.8 μg) or similar amount of BSA were added. At the top part of each panel, there is a schematic presentation of the substrate used in the experiments. The bottom part of panels A and B, and right part of panelC show quantification of the corresponding experiments. F, tetrahydrofuran; U, uracil. Values represent the mean ± sd of 3 independent experiments. **P < 0.01, ***P < 0.001 (2-sided Student’s t test).

We then asked what mechanism is responsible for the Ku70’s positive effect on BER process. Because Pol-β is a polymerase, our attention was first focused on whether the polymerase activity of Pol-β would change with Ku70 added into the reaction. The result showed that Ku70 itself has no polymerase activity and cannot extend the primer (Fig. 3C, lane 2). However, it can significantly promote Pol-β activity (Fig. 3C, lanes 4, 5, and 6), whereas addition of BSA has no such an effect (Fig. 3C, lanes 7, 8, and 9). In conclusion, our data indicate that Ku70 contributes to BER through enhancing the polymerase activity of Pol-β.

Knockdown Ku70 down-regulates SP-BER and LP-BER in vivo

Our observation that Ku70 can improve BER efficiency in reconstituted BER assays allowed us to propose that Ku70 not only plays key role in the NHEJ process but also has an important function in BER. To check this hypothesis, we used siRNA to knock down expression of Ku70 in HEK293T cells. Western blot data showed that the expression of Ku70 was successfully suppressed in Ku70 siRNA-transfected cells (Supplemental Fig. S4). Then, the whole-cell extracts were prepared to test BER efficiency. In this assay, the U-containing (Pol-β-U) and F-containing (Pol-β-F) unlabeled substrates were used for SP-BER and LP-BER, respectively (Fig. 4A, B and Supplemental Table S1). Incorporation of [³²P]-dCTP into single nucleotide gap generated by UDG and APE results in labeled 21mer unligated and 41mer fully repaired products (Fig. 4A, B). We observed that U or F lesions were efficiently repaired by cells treated with scramble siRNA but not by Ku70 knockdown cells (Fig. 4A, B). To demonstrate that Ku70 knockdown cells are impaired in BER, we carried out the SP-BER and LP-BER assay using cell lysates prepared from MMS-treated cells. The data showed that the efficiency of BER was dramatically decreased after Ku70 knockdown (Supplemental Fig. S5A, B). To exclude the possibility that Ku70 knockdown effects BER efficiency because of the reduced expression of BER proteins, we examine the protein levels of BER core components (Pol-β, APE1, FEN1, and PCNA). We found no effect of Ku70 knockdown on BER protein level (Supplemental Fig. S4), suggesting that the impact of Ku70 on the BER process is not caused by affecting the expression of BER proteins.

Figure 4.

Ku70 knockdown reduces BER efficiency. A, B) SP-BER and LP-BER were carried out with different concentrations of extracts prepared from Ku70-depleted cells or control cells as described in Materials and Methods. The top part of each panel shows the schematic structure of the corresponding DNA substrate. The bottom charts are a quantification of 3 independent repeats of the experiments for A and B, respectively. ***P < 0.001 (2-sided Student’s t test).

Ku70 knockdown cells are more sensitive to MMS

Stimulation of an in vitro BER reaction by Ku70 suggests that DNA end-joining proteins may be implicated in the cellular response to DNA damage repaired through BER pathway. Pol-β deficiency has been reported for impaired BER efficiency (31, 32); to test whether Ku70 has the similar effect, MMS sensitivity assay was performed in HEK293T cells transfected with Ku70 siRNA or scramble siRNA. As shown in Fig. 5A, B, Ku70 knockdown leads to increased cell mortality after MMS treatment. We also performed alkaline Comet assay to demonstrate accumulation of SSBs after MMS treatment of Ku70 knockdown cells. Ku70 siRNA–treated cells were obviously less effective in BER when compared with the control cells, showing a higher percentage of DNA in the tail after MMS treatment (Fig. 5C, D). The NADPH depletion assay (25) result also indicated that Ku70 is important for SSB removal (Supplemental Fig. S6). Taken together, these results indicate that Ku70 is involved in the processing of DNA lesions by BER in living cells.

Figure 5.

Suppressing Ku70 sensitizes cells to MMS and induces more DNA single-strand breaks. A) HEK293T cells transfected with Ku70 siRNA or scramble siRNA were captured under the microscope after 6-h treatment with 1 mM MMS. Scale bar, 30 μm. B) Relative cell viability was examined after 12-h treatment with 1 mM MMS. Statistical analysis of 3 independent experiments was accomplished as indicated. C, D) HEK293T cells transfected with Ku70 siRNA or scramble siRNA were treated with 1 mM MMS for 30 min, and percentage of DNA in the tails was quantified by the Comet assay immediately after treatment. Scale bar, 10 μm. Data represent the mean ± sd of 3 independent experiments. **P < 0.01, ***P < 0.01 (2-sided Student’s t test).

Pol-β participates in NHEJ

The Ku70 and Pol-β interaction was enhanced not only by MMS treatment but also by ETO stimulation, which prompts us to examine whether Pol-β functions in NHEJ. First, a well-characterized green fluorescent protein (GFP)-based reporter system, which has been previously established by Gunn and Stark (22) (see Materials and Methods for additional details), was used to measure the NHEJ efficiency (Fig. 6A, B). After transfection with scramble or Pol-β siRNA for 72 h, the reporter cells were collected and analyzed by flow cytometry. Interestingly, in Pol-β knockdown cells, the GFP-positive population was decreased by about 40% compared with the control cells (Fig. 6C), suggesting that Pol-β may play a role in cellular NHEJ. The same assay was also executed in Pol-β overexpressing cells; however, there was no significant change of total NHEJ efficacy (Supplemental Fig. S7), and this may be because the amount of endogenous Pol-β protein is enough to repair DNA breaks. There are 2 subpathways in NHEJ: c-NHEJ and alt-NHEJ; alt-NHEJ acts as a backup DSB repair pathway (14, 33). To determine the subpathway of NHEJ, which involves Pol-β, another reporter system described as alt-NHEJ-GFP (23) was used to examine the alt-NHEJ efficiency in Pol-β knockdown cells (Fig. 6D). Fluorescence-activated cell sorting (FACS) analysis data showed that Pol-β knockdown only slightly reduced (∼9%) GFP-positive cells compared with the control group (Fig. 6D). These results illustrate that Pol-β mainly stimulates c-NHEJ pathway. To demonstrate colocalization of Pol-β and phosphorylated H2AX (γ-H2AX) at the sites of DSB, we conducted immunofluorescence assay after treatment of cells with ETO. Cells were treated with DMSO or ETO and then stained with anti–Pol-β and γ-H2AX antibodies. As expected, the colocalization of Pol-β and γ-H2AX could be detected in ETO-treated cells (Fig. 6E). γ-H2AX is the variant of histone H2AX phosphorylated at Ser193 (34), and the phosphorylation of H2AX is one of the primary responses to the formation of DSBs. Therefore, the colocalization of Pol-β and γ-H2AX after ETO treatment further proves the role of Pol-β in DSB repair. Taken together, our results indicate that Pol-β participates in DSB repair by stimulating c-NHEJ pathway.

Figure 6.

Pol-β involves in NHEJ-mediated DNA repair. A, B) The schematic representation of EJ5-GFP (A) and alt-NHEJ-GFP (B) reporter system. C) U2OS EJ5-GFP cells were transfected with Pol-β shRNA or control shRNA for 24 h, and then cells were transfected with HA-I-SceI expression plasmid and continuously cultured for 48 h. Cells were collected for FACS analysis to identify GFP-positive cells. D) alt-NHEJ-EGFP U2OS cells were transfected with Pol-β shRNA or control shRNA for 24 h and then transfected with the HA-I-SceI expression plasmid and further incubated for 48 h; cells were then collected for FACS analysis. Knockdown efficiency and transfection efficiency were confirmed by Western blot as shown in the bottom part of panels C and D). NS, no significance. Data represent the mean ± sd of 3 independent experiments. **P < 0.01 (Student’s t test). E) Colocalization of Pol-β and γ-H2AX in DMSO or 10 μM ETO-treated HEK293T cells. Representative images are shown. Pol-β (green), γ-H2AX (red), or merged images are indicated. Scale bar, 10 μm.

Pol-β knockdown increases cell sensitivity to ETO and delays DSB repair

Previous studies have reported that Ku70 plays an important role in DSB repair (13–15). To further address whether Pol-β–deficient cells exhibit similar defects in rejoining DSBs under DNA damage, Pol-β knockdown stable cell line was generated by transfection with short hairpin (shRNA) targeting Pol-β. Knockdown efficiency was evaluated by Western blot (Fig. 7A). Because γ-H2AX is used as a standard marker for DNA DBS repair efficiency, we investigated whether Pol-β knockdown affected the level of γ-H2AX and its foci formation after ETO treatment. Obviously, both control and Pol-β knockdown cells appeared to be at the maximum levels of γ-H2AX at the 1.5-h time point after ETO treatment; however, Pol-β knockdown seemed to induce a significant increase in γ-H2AX levels compared with the control group (Fig. 7A, B). Moreover, the total level of γ-H2AX in Pol-β knockdown cells was continuously higher than in the control samples at all recovery time points (Fig. 7A, B). γ-H2AX foci formation assay presented in Fig. 7C, D fully confirmed the data from Western blot analysis. Similar experiments were performed with X-ray, another DSB inducer; although the maximum levels of γ-H2AX after irradiation appeared later after ETO treatment, the total level of γ-H2AX was consistently higher in Pol-β knockdown cells (Supplemental Fig. S8). To enforce the role of Pol-β in NHEJ repair of DSB, ETO sensitivity assay was performed in Pol-β–deficient and control HEK293T cells. We found that Pol-β knockdown leads to increased cell mortality after ETO treatment (Fig. 7E). Together, these data demonstrate that Pol-β suppression impairs cellular ability to repair DSB and leads to an increased amount of DNA damage foci and cell mortality upon ETO treatment.

Figure 7.

Pol-β–deficient cells exhibit more sensitive to ETO and delay DSB repair. A) Representative Western blot. Scramble shRNA and Pol-β shRNA–transfected HeLa cells were treated with DMSO or 10 μM ETO for 1.5 h and harvested at the indicated time points after treatment. Cells were lysed, and Western blot analysis was performed using specific antibodies against Pol-β, Ku70, γ-H2AX, Ku80, and Tubulin. B) Densitometric quantification of γ-H2AX from 3 independent experiments was performed with ImageJ software. C) Scramble shRNA and Pol-β shRNA–transfected HeLa cells were treated same as in A, and then fixed at the indicated times. Cells were stained with an anti–γ-H2AX antibody, and nucleus was stained using DAPI and then visualized by a fluorescence microscope. D) Densitometric quantification of γ-H2AX. Data represent the mean ± sd of 100 cells from 3 independent experiments. E) HeLa cells transfected with scramble shRNA or Pol-β shRNA were treated with indicated concentrations of ETO, and cell survival was analyzed. Data represent the mean ± sd of 3 independent experiments. Scale bar, 10 μm. **P < 0.01, ***P < 0.001 (2-sided Student’s t test).

HR is another important pathway responsible for the accuracy of DSB repair. To determine whether this pathway is also involved in repairing ETO-induced DSBs, we utilized the inhibitors of HR and c-NHEJ. NU7441 is a highly potent and specific inhibitor of c-NHEJ that targets DNA-PKcs, which is the key component of the c-NHEJ, whereas RI-1 is a RAD51-inhibitory compound that inactivates HR by directly binding to RAD51 protein surface (35). Therefore, HEK293T cells were pretreated with NU7441 or RI-1 for 14 h to inhibit NHEJ or HR, respectively. DMSO treatment was used in control experiments. Afterward, we measured γ-H2AX levels at the indicated time points after ETO treatment (Supplemental Fig. S9A, B). We found that NU7441 pretreatment triggers higher levels of γ-H2AX induced by ETO and delays dephosphorylation of γ-H2AX (Supplemental Fig. S9A, lanes 6, 7, and 8) compared with DMSO pretreatment (Supplemental Fig. S9A, lanes 2, 3, and 4). However, RI-1 pretreatment has no significant effect on the dynamic changes of γ-H2AX (Supplemental Fig. S9A, lanes 10, 11, and 12). Importantly, we also observed that NU7441 or RI-1 treatment without ETO stimulation did not cause a remarkable increase in γ-H2AX (Supplemental Fig. S9A, lanes 1, 5, and 9). These results suggest that c-NHEJ rather than HR is primarily responsible for ETO-induced DSB repair, suggesting the role of Pol-β in NHEJ rather than in HR. Altogether, our results provided clear evidence that Pol-β is involved in the efficient repair of DSBs through the c-NHEJ pathway and is essential for cell survival following ETO-induced DNA damage.

DISCUSSION

Pol-β and Ku70/80 are the key DNA repair proteins, which participate in BER and NHEJ, respectively. In the present study, Ku proteins were found after immunoprecipitation of cell extract with the Pol-β antibody, which suggests that in cells, at least part of the Ku heterodimer is in a complex with Pol-β. Although both Ku70 and Ku80 are found in the Pol-β precipitates, in vitro binding assays with purified proteins confirmed that only Ku70 directly interacts with Pol-β. It is worth mentioning that Ku-Pol-β complex detected in cells without exogenous DNA damage stimuli could be due to the endogenous DNA damage. We found that the association between Ku dimer and Pol-β increased under MMS or ETO treatment, suggesting that the complex is formed on DNA damage sites. This is consistent with our immunostaining data showing a higher degree of spatial colocalization between Ku70 and Pol-β in the nucleus following DNA damage by MMS or ETO. Because DNA damage induced by MMS and ETO is mainly repaired by BER or NHEJ respectively, in combination with our results, it seems reasonable to conclude that Pol-β and Ku70 interaction is necessary for both BER and NHEJ.

Previous works reveal that Pol-β plays a central role in BER, and a number of regulatory factors regulate BER by targeting Pol-β. For example, in LP-BER, strand displacement synthesis by Pol-β can be stimulated by FEN1 (36). Recent work has reported that HSP90 acts as a BER regulator by adjusting the interaction between X-ray repair cross-complementing gene-1 and Pol-β (37). Our results uncovered a new role of Ku70 as a regulatory factor for BER, promoting polymerase activity of Pol-β. Ku protein is well-known to work as a scaffold that recruits other repair factors of nonhomologous end joining and facilitates the following repair process. Interestingly, recent work from Roberts et al. (38) stated Ku as an effective 5′-dRP/AP lyase. Then, there may be a possibility that Ku70 could directly promote BER efficiency partly through its 5′-dRP/AP lyase activity. However, Ku is only active in removing dRP groups of the DSB ends, and it hardly works at this step at the single-strand breaks (38). Thus, it seems reasonable to conclude that Ku70 effect on BER is independent on its 5′-dRP/AP lyase activity.

In addition to protein-protein interaction, the post-translational modification also affects Pol-β activity and stability. It was reported that acetylation and arginine methylation of Pol-β is essential for its enzymatic activity, whereas ubiquitination is highly related to the protein stability (39, 40). Several E3 ubiquitin ligases and deubiquitinases are discovered to regulate BER by controlling the steady-state levels of Pol-β (40–42). It is worth mentioning that Ku70 may function as a deubiquitinase to regulate apoptosis through directly deubiquitinating Bax and Mcl-1 (43, 44). We then tested whether Ku70 executes its function in BER via deubiquitinating and stabilizing Pol-β or other BER proteins. We confirmed by Western blot analysis that the expression levels of BER proteins did not change after knockdown of Ku70. Therefore, these data exclude the possibility that Ku70 can regulate the stability of BER proteins through its deubiquitinase activity to promote BER efficiency.

In the NHEJ pathway, broken DNA ends are brought together, processed, and then ligated to restore DNA integrity. Before ligation, the DNA ends need to be processed by nuclease to remove noncomplementary nucleotides and fill in any remaining gaps. As mentioned earlier, 3 members of the Pol X family, including Pol-μ, Pol-λ, and TdT, are implicated in these reactions. Pol-μ associates with Ku in cell extracts and requires both Ku and XRCC4-ligase IV to form a stable complex on DNA in vitro. TdT is suggested to be involved in the repair of DSBs during V(D)J recombination, whereas Pol-λ efficiently fills in short gaps in DNA mimicking NHEJ. Additionally, another work has highlighted that Pol4 (Pol-β equivalent in yeast) is indeed involved in yeast NHEJ pathway and dependent on its nucleotide transferase function as well as its unique amino terminus. Mutation of Pol4 led to at most a 2-fold reduction in the frequency of joins that require only DNA polymerization (45).

In our study, we clearly showed that Pol-β depletion results in a 40% reduction of total NHEJ efficiency, which implies the role of Pol-β in NHEJ. This result is consistent with the recent view of Pol-β’s participation in end joining recently proposed by Ray et al. (46). Although they also observed the new function of Pol-β involvement in NHEJ, the details of the mechanisms are different. Ray et al. (46) reported that Pol-β participates in alt-NHEJ and that the ability to perform efficient alt-NHEJ is significantly reduced by Pol-β depletion, whereas our data showed that the reduction of NHEJ efficiency by Pol-β knockdown is mainly due to the c-NHEJ because the alt-NHEJ was only slightly affected by the same treatment. This discrepancy could be due to the different alternative end joining reporter systems. Ray et al. (46) employed a dual fluorescent protein reporter assay system to monitor both alt-NHEJ and HR at the same time. In that system, the EJ-RFP cassette was previously integrated into U2OS cells harboring the DR-GFP reporter for HR (47), whereas in our work, the alt-NHEJ-EGFP system harbors a full-length EGFP cassette that is interrupted by inserting a 27 bp oligonucleotide containing an I-SceI cleavage site, which is flanked by 9 nt of microhomology sequence (23). There are many assays to study NHEJ repair; however, the results sometimes are controversial regarding the dependencies of various NHEJ proteins on repair activity. For example, some reports stated that Ku70/80 down-regulation does not affect the overall NHEJ repair frequencies using an assay based on 2 I-SceI sites (48), whereas other studies suggest that depletion of these proteins has a significant effect on repair rates in similarly designed NHEJ assays (47, 49). Thus, we speculate that the reason for the different results may be due to the different detection methods for alt-NHEJ. Ku70/Ku80 complex efficiently binds to DSB as well as to nicked DNA (50). We propose that the major effect of the Ku70/Ku80 complex in BER is due to stimulation of Pol-β binding to the nicked DNA intermediate arising during BER (Fig. 8). The c-NHEJ process is initiated by binding the Ku70/80 heterodimer to both ends of the broken DNA molecules. Binding of Ku to DNA is a critical step that creates a scaffold to recruit other key enzymes necessary for c-NHEJ (51). Thus, similar to other Pol X family members involved in c-NHEJ, Pol-β may be recruited to DNA by Ku70 as a polymerase catalyzing gap-filling synthesis on the broken DNA ends in the NHEJ process.

Figure 8.

Graphical summary of the model for BER-NHEJ crosstalk mediated by Ku70 and Pol-β supported by this study. BER is the major SSB repair pathway, and c-NHEJ is the major DSB repair pathway. When SSBs occur, Ku70 binds to Pol-β and promotes polymerase activity of Pol-β, thus accelerating BER. Reciprocally, when DSBs occur, Pol-β binds to Ku70 and participates in c-NHEJ to promote repair of the DSB lesions.

To summarize, this study reveals one of the mechanisms of a crucial crosstalk between the NHEJ and BER pathways. It will be interesting to examine the role of NHEJ-Ku70 and BER-Pol-β interaction in tumorigenesis, in which both DNA repair pathways play important roles.

Supplementary Material

This article includes supplemental data. Please visit http://www.fasebj.org to obtain this information.

Glossary

Alt-NHEJ

alternative NHEJ

APE1

apurinic endonuclease 1

BAX

BCL2 associated X

BER

base excision repair

c-NHEJ

classical NHEJ

DNA-PK

DNA-dependent protein kinase

DNA-PKcs

catalytic subunit of DNA-PK

DSB

double-strand break

dRP

5’-deoxyribosephosphate lyase

ETO

etoposide

FACS

fluorescence-activated cell sorting

FAM

fluorescein

FEN1

flap endonuclease 1

GFP

green fluorescent protein

H2AX

H2A histone family member X

HR

homologous recombination

I-SceI

intron-encoded endonuclease I-SceI

Lig I

DNA ligase I

Lig IIIα

DNA ligase IIIα

LP-BER

long-patch BER

Mcl-1

myeloid cell leukemia 1

MMS

methyl methanesulfonate

NHEJ

nonhomologous end-joining

PCNA

proliferating cell nuclear antigen

Pol-β

DNA polymerase-β

shRNA

short hairpin RNA

siRNA

small interfering RNA

SP-BER

short-patch BER

SSB

single-strand break

TdT

terminal deoxynucleotidyl transferase

UDG

uracil-DNA glycosylase

XRCC4

X-ray repair cross complementing 4

γ-H2AX

phosphorylated H2AX

AUTHOR CONTRIBUTIONS

W. Xia and F. Pan designed the research; W. Xia and S. Ci performed the research; M. Li, M. Wang, L. Li, and K. Hua contributed new reagents or analytic tools; W. Xia and F. Pan analyzed data; W. Xia, Z. Ma, K. K. Alagamuthu, L. Qing, L. Luo, and L. Liu discussed and interpreted the data; and W. Xia, S. Ci, G. L. Dianov, Z. Hu, L. He, F. Pan, and Z. Guo wrote the manuscript.